|
 |
|
EMBO reports 8, 10, 887 (2007)
doi:10.1038/sj.embor.7401068
Comment on Frank Gannon's EMBO reports editorial "Address bias"
Rui Sousa
|
|
|
|
Rui Sousa is Professor of Biochemistry at the University of Texas Health Science Center in San Antonio, TX, USA.
e-mail: sousa@biochem.uthscsa.edu
|
|
|
|
It is a welcome sign that Frank Gannon discussed the problem of 'address bias' in his recent editorial (Gannon, 2007). Clearly, few active scientists could claim to be totally immune to this prejudice against lesser-known research centres, which subtly influences the review processes for publications, grants and fellowships. I remember once reading a letter to Nature from a researcher who, after a productive post-doc in Europe, had taken up a position in Brazil. In his new post, he was dismayed to discover that it was much harder to get his papers published, even though he felt they were equal to or better than anything he had published in England (Rumjanek, 1996). If we believe that science is best served by distributing resources and recognition on the basis of ability as demonstrated through productivity, rather than by following the prestige of address or pedigree, then we should agree that eliminating address bias would benefit science and humanity at large.
But how do we overcome address bias? One way would be to conceal an applicant's or author's identity and address—but this would not work for grant and fellowship applications, for which the track record, training and environment of the applicant are essential to the evaluation process. However, there is no convincing reason to reveal the addresses of manuscript authors to reviewers—indeed, reviewers remain anonymous, so why not also authors? Some might argue that such information is an important part of the review because it provides the reviewer with knowledge of the reliability and track record of a given lab. However, this seems more like an argument to conceal address bias in the reviewer.
Furthermore, the argument that the authors' address is important because often too little is presented in a manuscript to allow for a substantive review seems equally weak. The increasing trend to accompany published papers with ever greater amounts of supplementary data should make it possible to refute this argument by requiring that authors supply the information needed for a substantive review. This would probably enhance, not decrease, the critical nature of the review.
It is true that, even with the removal of the identities of the authors, not all papers would be truly anonymous. Referees might know that a particular study was about to come from a particular lab, or they might recognize an investigator's writing or research style. But even then, they would only be guessing. In most cases, however, the reviewers would have little idea of where a paper comes from—especially when the submitting laboratory is remote from major research centres—which would be exactly the point. More trivial issues, such as the removal of phrases like 'we previously showed', could be handled by instructing authors to avoid any explicit statements that reveal their identity. Such statements have always struck me as being too personal to be part of a scientific paper anyway and are usually factually untrue—the 'we' usually refers to the principal investigator's laboratory and the current authors on a paper are rarely identical to those on previous publications. Finally, whatever merit there might be in providing reviewers with the authors' identities—and there is some—it is overwhelmed by the benefits of eliminating address bias in the review process.
Given that Frank Gannon expresses concern regarding address bias, why not do something about it? Why not start a six-month to one-year-long experiment during which EMBO reports would withhold authors' names and addresses from the reviewers? If it works—although it is difficult to decide the criteria on which to measure success—and if other journals followed suit, it could start something of a revolution; it could genuinely democratize science by increasing the pace at which lesser-known research centres and universities gain recognition and resources. The risk seems minimal, especially if carried out as a limited experiment, and its results would be of interest to all scientists, on both sides of the review process. It would seem that the only reason to not undertake such an experiment would be that it risks antagonizing those well-established centres of research that benefit the most from address bias—which would rather prove the point.
|
|
References
Gannon F (2007) Address bias. EMBO Rep 8: 421 | Article | PubMed | ISI | ChemPort |
Rumjanek FD (1996) Paranoid about peer review? Nature 384: 509 | Article | PubMed | ISI | ChemPort |
|
|
top  |
|
|
|
